Rules of Expansion: an Updated Consensus Operator Site forthe CopR-CopY Family of Bacterial Copper Exporter SystemRepressors

Henrik O’Brien,a Joseph W. Alvin,a Sanjay V. Menghani,a Yamil Sanchez-Rosario,a Koenraad Van Doorslaer,a,b,c

Michael D. L. Johnsona,b,d

aDepartment of Immunobiology, University of Arizona, Tucson, Arizona, USAbBIO5 Institute, University of Arizona, Tucson, Arizona, USAcSchool of Animal and Comparative Biomedical Sciences, Cancer Biology Graduate Interdisciplinary Program, Genetics Graduate Interdisciplinary Program, University ofArizona Cancer Center, University of Arizona, Tucson, Arizona, USA

dValley Fever Center for Excellence, University of Arizona, Tucson, Arizona, USA

Henrik O’Brien and Joseph W. Alvin contributed equally to this work with order determined by a coin flip.

ABSTRACT Copper is broadly toxic to bacteria. As such, bacteria have evolved spe-cialized copper export systems (cop operons) often consisting of a DNA-binding/copper-responsive regulator (which can be a repressor or activator), a copper chap-erone, and a copper exporter. For those bacteria using DNA-binding copper repressors,few studies have examined the regulation of this operon regarding the operator DNAsequence needed for repressor binding. In Streptococcus pneumoniae (the pneumococ-cus), CopY is the copper repressor for the cop operon. Previously, homologs of pneumo-coccal CopY have been characterized to bind a 10-base consensus sequence T/GACANNTGTA known as the cop box. Using this motif, we sought to determine whethergenes outside the cop operon are also regulated by the CopY repressor, which waspreviously shown in Lactococcus lactis. We found that S. pneumoniae CopY did notbind to cop operators upstream of these candidate genes in vitro. During this pro-cess, we found that the cop box sequence is necessary but not sufficient for CopYbinding. Here, we propose an updated operator sequence for the S. pneumoniae copoperon to be ATTGACAAATGTAGAT binding CopY with a dissociation constant (Kd)of �28 nM. We demonstrate strong cross-species interaction between some CopYproteins and CopY operators, suggesting strong evolutionary conservation. Taken to-gether with our binding studies and bioinformatics data, we propose the consensusoperator RNYKACANNYGTMRNY for the bacterial CopR-CopY copper repressor ho-mologs.

IMPORTANCE Many Gram-positive bacteria respond to copper stress by upregu-lating a copper export system controlled by a copper-sensitive repressor, CopR-CopY. The previous operator sequence for this family of proteins had been iden-tified as TACANNTGTA. Here, using several recombinant proteins and mutationsin various DNA fragments, we define those 10 bases as necessary but not suffi-cient for binding and in doing so, refine the cop operon operator to the 16-basesequence RNYKACANNTGTMRNY. Due to the sheer number of repressors thathave been said to bind to the original 10 bases, including many antibiotic resis-tance repressors such as BlaI and MecI, we feel that this study highlights theneed to reexamine many of these sites of the past and use added stringency forverifying operators in the future.

KEYWORDS copper, metal, operator, operon, protein, repressor

Citation O’Brien H, Alvin JW, Menghani SV,Sanchez-Rosario Y, Van Doorslaer K, JohnsonMDL. 2020. Rules of expansion: an updatedconsensus operator site for the CopR-CopYfamily of bacterial copper exporter systemrepressors. mSphere 5:e00411-20. https://doi.org/10.1128/mSphere.00411-20.

Editor Craig D. Ellermeier, University of Iowa

Copyright © 2020 O’Brien et al. This is anopen-access article distributed under the termsof the Creative Commons Attribution 4.0International license.

Address correspondence to Michael D. L.Johnson, mdljohnson@arizona.edu.

Rules of expansion: an updatedconsensus operator site for the CopR/Y familyof bacterial copper exporter system repressorsvia @blacksciblog in @immunobiologyUA

Received 6 May 2020Accepted 7 May 2020Published

RESEARCH ARTICLEMolecular Biology and Physiology

crossm

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 1

27 May 2020

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

Metals are essential nutrients to all living organisms. Roughly 40% of proteins usemetal as cofactors and structural components in a vast number of cellular

processes (1). Iron, zinc, and manganese are examples of first-row divalent transitionmetals used by living organisms. The ability to form stable complexes plays a vital roleas to how each metal is used in the organism. The stability of protein-metal complexesis generalized by the Irving-Williams series (Mn � Fe � Co � Ni � Cu � Zn) (2). Ingeneral, more stable complexes correlate to a metal’s toxicity, as native metals for anactive site can be displaced by another metal ion further along in the observed series.This mismetallation is present across multiple metal-binding motifs and can result inabnormal protein function (3). For most prokaryotes, however, metals such as copper,nickel, and cobalt are broadly toxic. Higher order organisms have evolved ways toregulate and transport these metals, reducing promiscuity in mature active sites.Mammalian hosts have evolved strategies to both sequester essential metals frombacteria (e.g., Fe, Mn, and Ca) and bombard them with toxic metals such as copper; astrategy called nutritional immunity (4, 5). Mutations or deletions of a mammaliancopper transporter ATP7A can attenuate bacterial clearance by macrophages (6, 7).Thus, both within mammalian systems and on surfaces, copper is utilized as anantimicrobial (8–12).

Copper can catalyze Fenton-like reactive oxygen stress in bacteria, but mismetalla-tion of iron-sulfur clusters and enzymes necessary for nucleotide and amino acidsynthesis also leads to significant toxicity (13–17). As such, bacteria have evolvedspecialized import and export systems to acquire necessary metals and to adapt to thismetal toxicity. The presence of these import systems within the bacteria typicallyindicates a nutritional requirement for the metal. Iron, for instance, is an essential metalfor Streptococcus pneumoniae (the pneumococcus), a Gram-positive pathogen thatcauses pneumonia, meningitis, otitis media, and septicemia. In S. pneumoniae, iron hasfive known import systems/contributing genes, Pia, Piu, Pit, the hemin-binding systemencoded by SPD_1590 (D39 strain), and SPD_1609, but no known export systems(18–20). Calcium, zinc, and manganese all have export and import systems, whichsuggests that there are many optima of concentrations depending on the cell require-ments. However, the pneumococcus has no known import system for copper butcontains a dedicated copper export system encoded by the cop operon (19, 21–24). Ingeneral, copper export systems consist of an operon DNA regulator, a copper chaper-one, and one or two copper exporters (21, 25–31).

Multiple studies regarding the cop operon have been performed in S. pneumoniae(14, 21, 30, 32–35) and in other bacteria. Globally, cop operon regulators function aseither repressors, as in S. pneumoniae, or activators. Although there are cop operonactivators and repressors in structurally distinct groups, they all serve to protect thebacteria against copper stress by sensing copper and facilitating its export. Activators,such as CueR in Escherichia coli, sense copper and bind upstream of the cop operon topromote transcription (36). Conversely, repressor proteins release DNA upon bindingcopper and are found in species such as Lactococcus lactis and S. pneumoniae (CopR-CopY), Listeria monocytogenes, and Mycobacterium tuberculosis (CsoR) (19, 27, 30, 34, 35,37). The pneumococcal cop operon contains, copY as the DNA repressor, cupA as amembrane-associated copper chaperone, and copA as the copper-specific exporter (19,30, 34). The CopY repressor protein has an N-terminal helix-wing-helix motif, whichdirects the homodimer to bind its cognate DNA (34, 35). The Cu-chaperone proteinCupA chelates intracellular Cu, reduces it from Cu2� to Cu1�, and finally delivers it toCopA for export (32–34, 38). CupA copper chelation allows for the recycling of CopY tothe operator when copper is exhausted. Mutations in the copper export protein in copoperons result in decreased bacterial virulence, highlighting the importance of copperin nutritional immunity (21, 30, 39, 40).

Pneumococcal CopY is homologous to several known antibiotic resistance repres-sors, including BlaI, a Staphylococcus aureus MecI homolog that represses the gene fora �-lactamase (34, 35, 41). Like CopY, BlaI and MecI interact with a known operator

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 2

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

sequence, TACA/TGTA, form homodimers, and are mostly helical in secondary structure(34, 41). However, BlaI is regulated by proteases instead of metal interactions (42).

The CopR-CopY family of cop operon repressors have C-terminal metal-bindingprotein motifs, Cys-X-Cys. Each CXC motif binds copper in a 1:1 ratio. These motifs canalso bind zinc with a stoichiometry of two CXC motifs to one zinc (35). Since theseproteins function as dimers, the stoichiometry is one zinc per dimer, or two coppers perdimer after zinc displacement. Copper binding to CopR-CopY proteins causes a con-formational change, leading to release from the operator site, while zinc binding leadsto higher cop operator affinity (34, 35). However, how binding metal directly leads tothe conformational changes associated with DNA binding is currently unknown. Pre-vious work in Enterococcus hirae demonstrated that the CopY protein binds to twopalindromic regions (TACANNTGTA) upstream of the cop operon by using a DNase-footprinting assay (25). The cop operator sequence was then identified using surfaceplasmon resonance to evaluate binding of CopR-CopY to promoter regions in E. hirae,Lactococcus lactis, and Streptococcus mutans that contained TACANNTGTA (43). Theoligonucleotides used in this study contained a promoter sequence beyond just the 10bases listed above. We suspect that the 10 bases were identified as the operatorsequence based on the palindromic nature of the sequence, without testing whetherthose bases alone were sufficient for CopY binding.

Since the CopR protein from L. lactis was found to regulate multiple genes, wesought to test potential CopY-regulated genes outside the cop operon in Streptococcuspneumoniae (44). However, we quickly found that the TACANNTGTA operator is not areliable predictor for repressor protein binding (44). We used S. pneumoniae CopY inbinding experiments with various DNA constructs to define a sufficient cop operator invitro. Here, we present our findings and propose an update to the cop operator motif.We found that unlike the LacI repressor system, pneumococcal CopY does not dem-onstrate any cooperative binding, despite the operator sites being in close proximity.Furthermore, we highlight the strong conservation between bacterial cop operatorsand CopR-CopY constructs by demonstrating the presence and strength of cross-species interactions. We identified bacteria with CopY homologs and aligned theupstream sequences to determine if they also contained putative cop operators andwhether these operators followed the consensus motif we have proposed.

RESULTSCopY operator homology. Early DNA-binding studies were carried out using a

CopY homolog from Enterococcus hirae on the interactions with the cop operonoperator (25, 43, 45). We observed that there are two large repeats upstream of thepneumococcal cop operon that include a 10-base sequence important for CopY bind-ing. These motifs in S. pneumoniae differed slightly from those observed in E. hirae(Fig. 1A). Although the amino acid sequence of the E. hirae copper repressor and theupstream binding repeats are highly similar to those of S. pneumoniae (34, 35) andcontain the 10-base sequence, E. hirae operators upon initial observation lacked the

FIG 1 TIGR4 has two 21-base repeats containing the consensus CopY operators. (A) Aligned 21-basesequences for the two CopY operators. (B) TIGR4 SP_0727 promoter region sequence (containing both21-base repeats) aligned with the E. hirae ATCC strain 9790. Identical bases are underlined for therespective regions containing the operator.

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 3

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

extended regions flanking this sequence in pneumococcus (Fig. 1B) (25). A BLASTsearch revealed that the 61-base stretch of DNA upstream of the pneumococcal copoperon that includes the two extended 21-base repeats is highly conserved in allpneumococcal strains (46).

CopY binds to both pneumococcal cop operators. Previous studies showed thatCopR-CopY specifically bound to the cop operon operator in sequence- and metal-dependent manners, as disrupting the operator bases or adding copper disruptedbinding, while adding manganese or iron had no detectable effect (34). These studieswere conducted with only one full T/GACANNTGTA motif (here, KACANNTGTA) intact(34, 35). It is not clear why these operators are duplicated or why they fail to adhere tostrict palindrome sequences. We preliminarily sought to first determine if CopY boundto a DNA fragment containing both operators and if we could observe binding to oneor both operators. Using a 61-base double-stranded DNA (dsDNA) fragment andrecombinant CopY, we used an electrophoretic mobility shift assay (EMSA) to measurebinding. As expected, CopY binds in a dose-dependent manner to the 61-base dsDNAfragment (Fig. 2). Furthermore, there were two shifts displayed on the EMSA, indicatingthat the second-site interaction was concentration dependent (Fig. 2). These data arein agreement with initial studies of CopR in E. hirae, which also contains two operatorsites.

To quantitatively determine the affinities to the operators, we used a higherthroughput system to determine affinity: a biolayer interferometry (BLI) assay using theOctet Red384 (Octet). BLI is an optical technique similar to surface plasmon resonancethat uses interference of white light reflected from two surfaces to measure interactionsbetween molecules but is performed in a 96- or 384-well plate (47). The processinvolves (i) a baseline step in buffer; (ii) a step for the loading of analyte A to the surfaceof a biosensor; (iii) a second baseline step to make sure analyte A stays on the sensor(which should appear flat); (iv) an association step of analyte B (which should appear asa shift in reflected light if binding occurs); and (v) a disassociation step (which shouldappear as an opposite shift in the reflected light as analyte A disassociates from analyteB). In our experiments, analyte A was the biotinylated dsDNA oligonucleotides whichbind to the streptavidin biosensor, and analyte B was the recombinant protein (CopY).For viewing simplicity, we begin our plots begin at step 3. These experiments demon-strated that the two-operator-site (here, wild type) DNA had a similar affinity (dissoci-ation constant [Kd] � 28.1 nM) as that of the proximal DNA (DNA that has the distal sitescrambled) (Kd � 25.5 nM) to CopY (Table 1; Fig. 3A and C). The distal site DNA (DNAthat has the proximal site scrambled) had a slightly lower affinity (Kd � 55.2 nM)(Table 1; Fig. 3B). Since the sequences are identical, we suspect this difference isbecause the first 5= base in that motif is linked to biotin, in turn bound to thestreptavidin-coated biosensor. Thus, we would encourage including more than onebase between the DNA binding sequence and protein. As expected, CopY bound DNAconstructs containing intact 21-base repeat-containing cop operon operators withsignificantly higher affinities than for the scrambled DNA sequence (scram) (Fig. 3D) orsingle-stranded DNA (ssDNA) containing the wild-type operators, which both exhibited

FIG 2 CopY binds to both cop operon operators. EMSA with CopY and wild-type DNA. In seven of eightwells, a final concentration of 50 nM DNA was used with protein concentrations titrated by 2.5-folddilutions (640, 256, 102, and 41 nM). A final concentration of 640 nM CopY was used in a protein-without-DNA control with each replicate.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 4

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

extremely weak binding (data not shown). Taken together, CopY binds both 21-baserepeats containing the operator sites independently of each other with nanomolaraffinity.

Defining the pneumococcal CopY operator. With a reported binding sequence ofKACANNTGTA, we hypothesized that CopY may also bind similar sequences, as previ-ously seen in Lactococcus lactis (44). Allowing for one base variation from the reportedbinding sequence, we found matches upstream of genes upregulated under copperstress and hypothesized that they may also be regulated by CopY (14, 48). Sevenpotential binding sites were assessed using BLI. To our surprise, CopY did not bind tothese proposed operator sites (Fig. 4 and Table 2; see also Fig. S1 in the supplementalmaterial). Based on these results, we suspected that the reported sequence in theliterature may be necessary, but not sufficient, for CopY binding. Indeed, we found thatCopY failed to bind the reported binding sequence (Fig. 5A). Taken together, we have

TABLE 1 Data and model statistics from Octet kinetic experiments

Construct Kd (nM)a R2

Wild type 28.1 � 0.2 0.9474Distal site 55.2 � 0.4 0.9432Proximal site 25.5 � 0.2 0.947516 bp 360.0 � 3.0 0.980819 bp 37.1 � 0.3 0.955219 bp T to C 40.4 � 0.3 0.9408Full palindrome 39.6 � 0.4 0.9056E. hirae proximal 164.0 � 2.0 0.9269aThe Kd and model fit of dsDNA fragments containing wild-type or variations of the cop operator torecombinant S. pneumoniae CopY were assessed using BLI. At least 3 replicates were performed for eachconstruct.

FIG 3 Affinity measurements for CopY and the cop operon operators. DNA fragments were loaded onto a biosensor and tested with 1,000 nM (red), 500 nM(orange), 250 nM (yellow), 125 nM (green), 62.5 nM (blue), and 15.6 nM (purple) CopY wild type (A), distal site (B), proximal site (C), or scrambled (scram) (D).For each panel, data are representative of at least three experimental replicates.

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 5

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

concluded that the reported binding operator is necessary but not sufficient for coprepressor binding (25).

We next wanted to establish which bases outside the previously reported 10-baseKACANNTGTA sequence were necessary for CopY binding. Beginning with the minimal10-base sequence, constructs for BLI were designed to include or exclude identicalbases from the 21-base S. pneumoniae CopY operator. The first set of sequencesextended the operator five bases upstream or six bases downstream of the included10-base sequence. The remaining positions were left scrambled. Neither of thesefragments bound S. pneumoniae CopY more than the negative controls, suggestingthere are essential bases on each side of the 10-base sequence (Fig. 5B). We thenwidened the operator construct by two or three bases on either side of the core

FIG 4 Prediction of CopY binding based on 10-base sequence overestimates binding sites. CopY at 3 �Mwas used to assess binding to DNA fragments containing potential CopY operators upstream of therespective genes SP_0090 1 (green), SP_0090 2 (gray), SP_0045 (orange), SP_0530 (light green), SP_1433(red), with controls for wild type (blue), scrambled (scram; light blue), and no DNA (purple). Data arerepresentative of three experimental replicates.

TABLE 2 Outcomes of CopR or CopY binding to potential operator sites from L. lactis (44) or S. pneumoniae, respectively

Organism Color in Fig. 4Closest downstreamgene Sequencea

CopY-CopRbinding Reference

Lactococcus lactis NAb ytjD1 AAATAGTTTACAAGTGTAAATTTATTT Yes 44NA ydiD AAAATGTTTACATGTGTAAATTTTCAC Yes 44NA copR TTAGTGTTTACACGTGTAAACTTATCT Yes 44NA copB TGATAGTTTACAATTGTAAACTATATA Yes 44NA yahC TTTTCGTTTACAATTGTAAACATAGAA Yes 44NA lctO CTATCATCTACAGATGTAAACTTTATA Yes 44NA ytjD2 GATAAGATTACATATGTAAACAATAAA Yes 44NA yfhF TAAGTATATACATCTGTAAAACTGAAA No 44NA yxdE TTTGCTATTACACTTGTATCACATAAA No 44

Streptococcus pneumoniae Dark green Sp_0090 1 TGATTTAGGACATTTGTTTGATAGTGG No This studyGray Sp_0090 2 GAGTATACTAATAATGTAATCGTTATC No This studyOrange Sp_0045 GGTGAACTAACAGATGTTTACGAAATT No This studyLight green Sp_0530 ATTTGAGGAACAAATGTACGTTTATAA No This studyRed Sp_1433 GTAATTATAACAGATGTATAATAGAAA No This studyNA Sp_1863 ATGAATAAAACAATTGTAACACTCATC No This studyNA Sp_2073 AAGGCGGAAACATGTGTCAATGACTTG No This studyNA CopY (proximal site) GTGTAATTGACAAATGTAGATTTTGGA Yes This studyNA CopY (distal site) CTATAATTGACAAATGTAGATTTTAAG Yes This study

aUnderlined bases indicate bases varying from the reported 10-base consensus sequence (bold font).bNA, not applicable.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 6

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

10-bases for a total of 14 and 16 bases of the total 21 bases. These operators were usedto determine the minimum operator size for S. pneumoniae CopY. The 14-base se-quence did not display binding greater than that of negative controls (Kd � 3 �M), butthe 16-base sequence displayed approximately �10-fold weaker affinity than theproposed “full-length” 21-base operator. (Fig. 5C and Table 1; see Fig. S2). Extending thesequence to 19 of the 21 bases led to comparable levels of binding to the full sequence(Fig. 5D). Based on these data, we suggest that 16 bases make up a minimal operatorsite (ATTGACAAATGTAGAT) recognized by CopY and that additional flanking basesincrease stability.

Bioinformatic characterization of the CopY-CopR operator. We performedBLASTp searches for S. pneumoniae TIGR4 CopY homologs, first excluding and thenincluding the Streptococcus genus (46). Using a maximum target sequence number of1,000 for each search, and then combining both lists, we found 335 candidate ho-mologs (see Data Set S1, Tab 1). From this list, we extracted protein sequences from theunique NCBI accession numbers—many NCBI accession numbers represented severaland/or identical homologs. Many of the 141 unique protein sequences belonged tospecies in the Streptococcus and Lactobacillus genera (Data Set S1, Tab 2). This table alsoincluded species such as the yogurt probiotic Lactobacillus acidophilus and Mycobac-teroides abscessus, an emerging multidrug-resistant pathogen that causes lung, skin,and soft tissue infections (49).

FIG 5 The minimum operator for sufficient CopY binding is 16 bases in length. (A) Binding of the 10-base fragment compared to positive and negative controls:wild type (blue), Wild type no protein (red), 10-base sequence (gray), scrambled (scram; light blue), and no DNA (purple). (B) DNA fragments containing eitherthe upstream (5 bases) or downstream (6 bases) of the 10-base sequence within 21-base repeat compared to positive and negative controls: proximal site(orange), five bases upstream (pink), six bases downstream (light green), 10-base sequence (gray). (C) Fragments were used to assess extended sequence onboth sides of the 10-base sequence: proximal site (orange), 14 bases (black), 16 bases (yellow), 10 bases (gray), scram (light blue), and no DNA (purple). Fordata in panels A to C, we used 3 �M CopY to assess binding to each of the fragments. (D) A fragment containing 19 of the 21 bases in the repeat had comparablelevels of binding to the full repeat. The following concentrations were used to establish the Kd: 1,000 nM (red), 500 nM (orange), 250 nM (yellow), 125 nM (green),62.5 nM (blue), and 31.3 nM (purple) CopY. Data are representative of three experimental replicates.

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 7

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

We used programs within the MEME suite to identify DNA motifs within 100-basesupstream of the copY homolog start codon, which would correspond to the promoterand suspected operator-containing elements (Data Set S1, Tab 3) (50, 51). Similar tothat of known CopR-CopY operators, the program-derived 21-base sequence containedKACANNTGTA (Fig. 6A) (52). Of these 88 sequences, 67 had two CopR-CopY cop operonoperators (similar to those in Streptococcus pyogenes, E. hirae, and Enterococcus fae-cium), 14 had one operator (such as Enterococcus faecalis), and 7 lacked detectablesequence similarity (Fig. 6B to D; Data Set S1, Tab 4 to 6). Increasing the maximumbases upstream to 500 did not yield additional sequences (data not shown). For thegenomes with two operators, we found that most of the operators were between 24and 39 bases from each other, with the mode being 26 bases (Fig. S2). Of the speciesthat had a single cop operon operator, the sequences were more variable KACANNYGTA (Data Set S1, Tab 5). Of these 14 sequences, 7 were on the positive strand (butsix were palindromic), and 7 nonpalindromic sequences were on the negative strand(Data Set S1, Tab 5). Of the seven sequences that had no apparent operator homology,six were in the Lactobacillus genus and one was in the Macrococcus genus (Data Set S1,Tab 6). Comparing the upstream regions across these seven strains, MEME suite did notidentify a consensus motif. This suggests that these seven strains lost the operator sincediverging from a common ancestor.

Determining the consensus cop operon operator. The initial characterization ofthe CopY operator sites was performed using a few mutations in the E. hirae coppromoter (25, 53). As a proof of principle, we used the native E. hirae cop distal orproximal operators and examined them for pneumococcal CopY binding (Fig. 7A andB). S. pneumoniae CopY bound to the proximal E. hirae operator with low nanomolaraffinity (Fig. 7A). However, pneumococcal CopY did not bind to the distal E. hiraeoperator (Fig. 7B). This implies that for CopY from S. pneumoniae, the central “AA” motif(previously notated as “NN”) cannot accommodate variations at this position (Fig. 7B).

While inspecting the alignment of bacterial species operator sites via MEME Suite(Fig. 1A and 6A; Data Set S1, Tab 4), we identified a clear purine-N-pyrimidine patternon each side of the previous 10-base operator. Using our sequence data of homologouscop operators, we arrived at a proposed consensus sequence of RNYKACANNTGTARNY(where “R” is purine, “Y” is pyrimidine, and “K” is either G or T) (44) for CopY family

FIG 6 CopY cop operon operator consensus sequences in the genomic DNA. (A) Top consensus DNA sequence contained within the87 unique upstream 100 bases sequences to the respective CopY organism as detected by MEME suite. The literature-basedT/GACANNTGTA is underlined within the program-generated consensus sequence. (B) Total unique DNA sequences from Data Set S1,Tab 4, listed by number of CopY operators. (C) Unique genera with two CopY cop operon operators from Data Set S1, Tab 4. (D) Uniquegenera with one CopY operator from Data Set S1, Tab 5.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 8

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

repressor operators. These data were supported by both the E. hirae cop operatorshaving the alternate purine or pyrimidine bases at positions 1, 14, and 16 but are stillfunctional (Fig. 1B). Thus, we focused on position 3, which was the same in thepneumococcal operator and E. hirae operators in order to test our hypothesis thatconserved R-N-Y motifs flank the 10-base core at positions 1, 3, 14, and 16. Position 3was mutated to the alternate pyrimidine (T to C) or switched from pyrimidine to purine(T to A). As expected from our model, the T to C mutation at position 3 did notsignificantly alter the binding affinity compared to that of the 19-base fragment, whilethe T to A mutation completely abolished binding (Fig. 7C and D).

Due to the palindromic nature of this operator, we decided to test the last nonpal-indromic base in the now RNYKACAAATGTARNY consensus which was the “A” inposition 13. With position 4 being K (G or T), we mutated the “A” at position 13 to a “C”so that would fit with the complementarity compared to position 4, termed “fullpalindrome.” We found that pneumococcal CopY was able to bind to this full palin-drome sequence at near WT levels (Table 1). Taken together, we believe that thepneumococcal DNA binding sequence for CopY is RNYKACAAATGTMRNY, with “M”representing “C” or “A,” as opposed to the previously reported KACANNTGTA (Fig. 8A).

Interspecies compatibility in vitro across CopY family proteins and operators.Thus far, we have defined the pneumococcal CopY operator to be larger and moreselective than the canonical cop box motif. However, our bioinformatics data indicatethat there is strong conservation across the protein family operators (Fig. 8B). Wehypothesized that some degree of cross-species interactions or substitution could bepossible in vitro. Therefore, we produced recombinant CopY from both E. hirae and

FIG 7 Pneumococcal CopY binds to E. hirae DNA in accordance with the newly proposed consensus cop operon operator. Affinity of CopY binding to variousDNA fragments was determined using the following concentrations of CopY: 1,000 nM (red), 500 nM (orange), 250 nM (yellow), 125 nM (green), 62.5 nM (blue),and 31.3 nM (purple). (A) E. hirae proximal site. (B). E. hirae distal site (C) The 19-base fragment with a T-to-C mutation. (D) The 19-base fragment with a T-to-Amutation. For each panel, data are representative of three experimental replicates.

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 9

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

Streptococcus thermophilus to further test our proposed operator consensus and toglean details on how the proteins bind to their operators. We encountered problemswith initial purification of E. hirae CopY. Incubation with tobacco etch virus (TEV)protease led to proteolysis. The ideal sequence for TEV protease cleavage is ENLYFQ |GS, with the pipe character representing the hydrolyzed bond. Since this site is sospecific, it is commonly used for precision proteolysis, e.g., removing a purification tag.Reviewing the amino acid sequence of E. hirae CopY, we identified a functional TEV site(ENLFSH | IC). After reviewing the literature, we introduced an E86D mutation (DNLFSHIC) (54). This point mutation was sufficient to abrogate the undesired proteolysis whileretaining the negative charge at position 86.

We used BLI to measure binding between the S. pneumoniae, S. thermophilus, andE. hirae cop repressor proteins and CopY operator sites from S. pneumoniae, S. thermo-philus (proximal and distal), and E. hirae (proximal and distal). These experimentscomprised a small matrix of interactions between related CopY proteins and operators.The E. hirae cop repressor bound to all five operators with no detectable difference evenbetween uncleaved E. hirae CopY and cleaved E86D protein, implying that the His tagand TEV cleavage site intact had no effect on binding this specific system. We foundthat S. pneumoniae CopY did not bind to the distal E. hirae operator but did bind to theE. hirae proximal operator and both distal and proximal cop operators from S. thermo-philus (Tables 1 and 3). The S. thermophilus cop operon repressor bound to its ownoperators and the E. hirae operators (albeit weakly to the distal site) but not to thepneumococcal operator (Table 3). Taken together, the consensus sequence for theseoperators would be RNYKACANNTGTMRNY.

DISCUSSION

Previous binding studies characterizing pneumococcal CopY were carried out withDNA containing only one operator (21, 34, 35). Here, we demonstrate that CopR-CopYhomologs in Streptococcus and many other genera have two operators upstream of thecop operon (Fig. 1A). Despite having two identical 21-base repeat operators withinclose proximity, the S. pneumoniae CopY protein does not have a significantly higheraffinity for DNA with two operators present. A limitation of this study and subject tofuture direction is how these two operators affect gene regulation. Based on predicted

FIG 8 Chart representing the previous, newly proposed pneumococcal, and newly proposed CopR-CopYprotein family consensus cop operon operator. (A) Bases that change from the initial 10-base consensusoperator in the now total consensus operator are highlighted in red. (B) Sequence alignment of the S.pneumoniae, S. thermophilus, and E. hirae operator sites.

TABLE 3 Cross-species binding of CopY in vitro

CopY protein

Bindinga

E. hirae

S. pneumoniae

S. thermophilus

Distal Proximal Distal Proximal

E. hirae �� �� �� �� ��S. pneumoniae � �� �� �� ��S. thermophilus � �� � �� ��

aCopY from different bacteria was assessed for its ability to bind the operators found in other bacterialspecies to produce a binding matrix. Interactions were scaled based on relative affinities of binding. �, nobinding; �, weak binding; �� strong binding.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 10

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

�35 and �10 sites, we suggest that the distal operator prevents sigma factors frombinding at the copY �35 element and that the proximal operator occludes RNApolymerase, establishing two layers of repression for the cop operon (55). This reason-ing is also consistent with the proposed hypothesis of why there are two operators inthe antibiotic resistance repressor BlaI (56). It is unclear which sigma factor drives copYtranscription or whether the binding between CopY, a sigma factor, and/or RNApolymerase is competitive. Another subject of future direction is why such repressionthrough multiple operators is necessary for the cop operon especially considering that(i) cop operon upregulation is linked to increased pneumococcal survival in the hostand (ii) a ΔcopY mutant has increased virulence in mice (30, 57). We anticipate thatfurther study of these systems via fitness testing and exploring the effect of havingthese systems on under metal-limiting conditions will yield clues as to the competitiveadvantages or selective pressures of the cop operon at the host-microbe interface.

Given the previous 10-base consensus sequence suggested for CopR-CopY proteinoperators, it was plausible there were additional genes that the cop repressor couldregulate, similarly to L. lactis (44). Some of these putative binding sites in S. pneumoniaecorresponded to putative promoter regions of genes and operons upregulated undercopper stress, suggesting that CopY was a regulator of several operons (14). However,we have empirically demonstrated that these putative sites— despite showing copper-dependent regulation— did not bind CopY and thus, importantly, that the canonical10-base cop box is not sufficient for binding (14, 21). These facts ultimately led us toexplore and propose an expanded CopY operator sequence (RNYKACAAATGTMRNY),adding to the previously reported KACANNTGTA (Fig. 8). Unlike in the L. lactis genome,a search using this operator motif in multiple streptococcal species for potentialoperators yielded no additional sites (48).

While CopY does not bind to other locations within S. pneumoniae, its aforementionedrole in controlling the cop operon and the importance of that operon at the host-pathogeninterface makes it of continued interest. It is still not known how CopY has higher bindingto DNA with zinc bound than in the apo form and how copper changes the structure todisrupt DNA binding. This expanded operator can better inform structure/function studiesregarding residue-to-nucleotide interaction in the CopR-CopY family of proteins, for whichthere is still no structure with metal or DNA bound.

In generating the list of homologous CopR-CopY genes and proteins, we also wereable to look upstream of the gene and compile what we believe to be a consensusoperator site which differs slightly from the proposed pneumococcal operator. Thestrong conservation between repressor and operator across species is made clear bythe affinity studies with cop repressors from both E. hirae and S. thermophilus. However,this cross-species interaction was not universal—pneumococcal CopY only bound toone E. hirae operator but bound both S. thermophilus sites—the new consensusoperator is more predictive of these proteins binding in this limited homology matrix.Based on the strong interactions we observed in vitro, the selectivity of repressors foroperators, and a suite of predicted CopR-CopY operators, we propose a consensusoperator of RNYKACANNYGTMRNY (Fig. 8). This consensus operator is consistent withthe results presented by Magnani et al. using the L. lactis CopR protein (Table 2) (44).

We propose that, moving forward, operator sequences for transcriptional regulatorsshould be revisited and examined with a higher level of scrutiny, specifically those thatare just apparent palindromic sequences. In doing so, we believe that this would revealpotential binding interactions within the organism’s respective genome and withgenomes of the cohabitant bacteria of their environments from which they mightacquire new genes.

MATERIALS AND METHODSAligning and comparing CopY homologs and promoter sequences. The BLAST sequence align-

ment algorithm was used to align both E. hirae and TIGR4 S. pneumoniae cop operon promoter regions,the 21-base repeats upstream of the TIGR4 pneumococcal cop operon, and the promoter regions ofpneumococcal species (46). A set of custom Python scripts (available from https://github.com/Van-Doorslaer/O-Brien_et_al_2019) were used to assign identified copY homologs to bacterial genomes and

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 11

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

extract the suspected regulatory region from individual species (100 and 500 bases upstream of the startcodon). Importantly, in many cases, the initial BLAST search identified CopY homologs which matchedmultiple species isolates/strains. In this case, the identified proteins were again compared to the NCBIdatabase, and the homolog with the lowest E value was retained. If this approach was unsuccessful, thehomolog was excluded from further analysis.

Cloning and site-directed mutagenesis. copR-copY from S. pneumoniae, S. thermophilus, and E.hirae were cloned from genomic DNA of the respective organisms and inserted into the pMCSG7 vector(58). The PrimerX tool (http://bioinformatics.org/primerx) was used to generate primers for site-directedmutagenesis of E86D in the E. hirae copR gene. Recombinant Pyrococcus furiosus (Pfu) polymerasewas used to perform all molecular reactions. Correct sequences were verified by Sanger sequencingbefore use.

Protein purification. Recombinant CopR-CopY proteins were purified as described by Neubert et al.(34), with modifications. The pMCSG7 vector includes an N-terminal 6His tag linked to CopR-CopY viaa tobacco etch virus (TEV) protease cleavage site (58). Unless otherwise specified, all steps wereperformed on ice or at 4°C. After initial purification using immobilized metal-affinity chromatography(IMAC) (HisTrap FF; GE Healthcare), the crude CopR-CopY sample was incubated at 23°C with a 100:1mass ratio of recombinant TEV. The cleaved product was confirmed via SDS-PAGE. Cleaved CopR-CopYwas purified with subtractive IMAC; the TEV protease contains a C-terminal His tag. The flowthrough wasconcentrated using concentrator with a molecular weight cutoff (MWCO) of 10 kDa (MilliporeSigma) andfurther purified by size exclusion chromatography (SEC) (Superdex 200; GE Healthcare) using a buffer of20 mM Tris (pH 8), 200 mM NaCl, and 1 mM Tris(2-carboxyethyl)phosphine (TCEP). Based on the elutionvolume and comparison to an SEC standard (Bio-Rad), we observed that CopR-CopY proteins elutedexclusively as a dimer. Peaks containing pure CopR-CopY (as determined by SDS-PAGE) were pooled, andthe concentration was determined by absorbance at 280 nm. Samples were used immediately orprotected with 35% to 50% glycerol and aliquoted into thin-walled PCR tubes and flash-frozen usingliquid N2.

Electromobility shift assay. Primers for binding, 5=-TAATTGACAAATGTAGATTTTAAGAGTATACTGATGAGTGTAATTGACAAATGTAGATTTT-3= and 5=-AAAATCTACATTTGTCAATTACACTCATCAGTATACTCTTAAAATCTACATTTGTCAATTA-3=, were annealed by heating a solution containing 1:1 molar equivalents ofeach strand to 95°C and then reducing the temperature by �1°C/min to 22°C. EMSA buffer wasTris-borate (TB) electrophoresis buffer (EDTA was omitted to diminish metal chelation). Samples wereincubated at 4°C for 5 min, loaded onto a 5% polyacrylamide Tris-borate-EDTA (TBE) gel (Bio-Rad) thathad been prerun for 15 min in TB buffer. Samples were electrophoresed at 40 V for 120 min. Thepolyacrylamide gel was stained with 0.02% ethidium bromide (Amresco) and imaged with a Gel Doc XR�

system (Bio-Rad).Biolayer interferometry DNA/protein binding. Double-stranded DNA fragments were hybridized

as described above by incubating 5= biotinylated ssDNA with an unlabeled complementary strand at95°C for 5 to 10 min and then left to cool to room temperature (see Table S1 in the supplementalmaterial). The biotinylated dsDNA oligonucleotides were diluted to 250 or 50 nM in the assay buffer(50 mM Tris [pH 7.4], 150 mM NaCl, 4% glycerol, 1 mM TCEP, 0.1% bovine serum albumin [BSA]). Duringinitial optimization, we observed significant nonspecific binding of CopY to the biosensors in the absenceof BSA. The inclusion of 0.1% BSA eliminated signals of nonspecific CopY binding at and above thehighest concentrations used in the assay (3 �M). The dissociation constant for CopY with various DNAfragments was determined using an Octet Red384 (Pall ForteBio). Streptavidin biosensors (Pall ForteBio)were hydrated at 26°C using the Sidekick shaker accessory for 10 min at 1,000 rpm. Hydrated sensorswere incubated in the assay buffer to acquire a primary baseline. The sensors were then loaded withbiotinylated dsDNA, followed by a secondary baseline measurement using wells with buffer solution.DNA-loaded biosensors were then moved to wells containing various CopY concentrations to measurethe association and then placed back into assay buffer for dissociation recordings. All experiments weremaintained at 26°C with shaking at 1,000 rpm. The optimized protocol was as follows: primary baseline,60 s; DNA loading, 150 s; secondary baseline, 180 s; association, 180 s; and dissociation, 270 s. Themethods were optimized to smooth the signal at association and minimize recording at equilibria.Additionally, proteins were tested both with and without cleaving the N-terminal His tag.

Analysis was performed using Octet software. We applied a 1:1 binding model using a global fit tobiosensor replicates at each concentration of CopY. During preprocessing, an average of the secondarybaseline across the various biosensors was applied, as well as Savitzky-Golay filtering to reduce noise. Thedata were interstep corrected using an alignment to the dissociation step. Data were modeled usingcombined fits of absorption rate constant (ka) and dissociation rate (kd) values across independentreplicates. Final estimates for Kd and related statistics were taken from the kinetic analysis.

Rate constants for each sample were determined using the Octet analysis software as follows. For all1:1 stoichiometric modeling, complex formation was evaluated as pseudofirst-order kinetics. The ob-served rate constant (kobs) was calculated according to the equation Y � Y0 � A�1 � e–kobs�t�. Where Y0

is the initial binding, Y is the level of binding, t is time, and A is asymptote value at max response.Dissociation rate (kd) was calculated according to the equation Y � Y0 � Ae–kd�t. The calculated kobs

and kd values were then used to determine ka using the equation ka �kobs�kd

�CopY�. Finally, the

dissociation constant (Kd) was determined by the identity Kd �kd

ka.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 12

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

SUPPLEMENTAL MATERIALSupplemental material is available online only.FIG S1, PDF file, 0.2 MB.FIG S2, PDF file, 0.1 MB.FIG S3, PDF file, 0.3 MB.TABLE S1, PDF file, 0.1 MB.DATA SET S1, XLSX file, 0.1 MB.

ACKNOWLEDGMENTSWe thank Rachel Wong, the University of Arizona Functional Genomics Core Facility,

and Richard Yip for assistance and support using the Octet Red384.This work was funded by an NIGMS grant 1R35128653 (to M.D.L.J.).We declare no conflicts of interest.

REFERENCES1. Waldron KJ, Rutherford JC, Ford D, Robinson NJ. 2009. Metalloproteins and

metal sensing. Nature 460:823–830. https://doi.org/10.1038/nature08300.2. Irving H, Williams R. 1953. 637. The stability of transition-metal complexes.

J Chem Soc 1953:3192–3210. https://doi.org/10.1039/jr9530003192.3. Imlay JA. 2014. The mismetallation of enzymes during oxidative stress. J

Biol Chem 289:28121–28128. https://doi.org/10.1074/jbc.R114.588814.4. Kehl-Fie TE, Skaar EP. 2010. Nutritional immunity beyond iron: a role for

manganese and zinc. Curr Opin Chem Biol 14:218 –224. https://doi.org/10.1016/j.cbpa.2009.11.008.

5. Hood MI, Skaar EP. 2012. Nutritional immunity: transition metals at thepathogen-host interface. Nat Rev Microbiol 10:525–537. https://doi.org/10.1038/nrmicro2836.

6. Ladomersky E, Khan A, Shanbhag V, Cavet JS, Chan J, Weisman GA, PetrisMJ. 2017. Host and pathogen copper-transporting P-type ATPases func-tion antagonistically during Salmonella infection. Infect Immun 85:e00351-17. https://doi.org/10.1128/IAI.00351-17.

7. White C, Lee J, Kambe T, Fritsche K, Petris MJ. 2009. A role for the ATP7Acopper-transporting ATPase in macrophage bactericidal activity. J BiolChem 284:33949 –33956. https://doi.org/10.1074/jbc.M109.070201.

8. Braymer JJ, Giedroc DP. 2014. Recent developments in copper and zinchomeostasis in bacterial pathogens. Curr Opin Chem Biol 19:59 – 66.https://doi.org/10.1016/j.cbpa.2013.12.021.

9. Salgado CD, Sepkowitz KA, John JF, Cantey JR, Attaway HH, Freeman KD,Sharpe PAM, Michels HT, Schmidt MG. 2013. Copper surfaces reduce therate of healthcare-acquired infections in the intensive care unit. InfectControl Hosp Epidemiol 34:479 – 486. https://doi.org/10.1086/670207.

10. Warnes SL, Keevil CW. 2016. Lack of involvement of Fenton chemistry indeath of methicillin-resistant and methicillin-sensitive strains of Staphy-lococcus aureus and destruction of their genomes on wet or dry copperalloy surfaces. Appl Environ Microbiol 82:2132–2136. https://doi.org/10.1128/AEM.03861-15.

11. Macomber L, Hausinger RP. 2011. Mechanisms of nickel toxicity in micro-organisms. Metallomics 3:1153–1162. https://doi.org/10.1039/c1mt00063b.

12. Kumar V, Mishra RK, Kaur G, Dutta D. 2017. Cobalt and nickel impair DNAmetabolism by the oxidative stress independent pathway. Metallomics9:1596 –1609. https://doi.org/10.1039/c7mt00231a.

13. Macomber L, Imlay JA. 2009. The iron-sulfur clusters of dehydratases areprimary intracellular targets of copper toxicity. Proc Natl Acad Sci U S A106:8344 – 8349. https://doi.org/10.1073/pnas.0812808106.

14. Johnson MD, Kehl-Fie TE, Rosch JW. 2015. Copper intoxication inhibitsaerobic nucleotide synthesis in Streptococcus pneumoniae. Metallomics7:786 –794. https://doi.org/10.1039/c5mt00011d.

15. Djoko KY, Phan MD, Peters KM, Walker MJ, Schembri MA, McEwan AG.2017. Interplay between tolerance mechanisms to copper and acidstress in Escherichia coli. Proc Natl Acad Sci U S A 114:6818 – 6823.https://doi.org/10.1073/pnas.1620232114.

16. Djoko KY, McEwan AG. 2013. Antimicrobial action of copper is amplifiedvia inhibition of heme biosynthesis. ACS Chem Biol 8:2217–2223. https://doi.org/10.1021/cb4002443.

17. Macomber L, Rensing C, Imlay JA. 2007. Intracellular copper does notcatalyze the formation of oxidative DNA damage in Escherichia coli. JBacteriol 189:1616 –1626. https://doi.org/10.1128/JB.01357-06.

18. Miao X, He J, Zhang L, Zhao X, Ge R, He QY, Sun X. 2018. A novel iron

transporter SPD_1590 in Streptococcus pneumoniae contributing to bac-terial virulence properties. Front Microbiol 9:1624. https://doi.org/10.3389/fmicb.2018.01624.

19. Honsa ES, Johnson MD, Rosch JW. 2013. The roles of transition metals inthe physiology and pathogenesis of Streptococcus pneumoniae. FrontCell Infect Microbiol 3:92. https://doi.org/10.3389/fcimb.2013.00092.

20. Yang XY, He K, Du G, Wu X, Yu G, Pan Y, Zhang G, Sun X, He QY. 2016.Integrated translatomics with proteomics to identify novel iron-transporting proteins in Streptococcus pneumoniae. Front Microbiol 7:78.https://doi.org/10.3389/fmicb.2016.00078.

21. Shafeeq S, Yesilkaya H, Kloosterman TG, Narayanan G, Wandel M, An-drew PW, Kuipers OP, Morrissey JA. 2011. The cop operon is required forcopper homeostasis and contributes to virulence in Streptococcus pneu-moniae. Mol Microbiol 81:1255–1270. https://doi.org/10.1111/j.1365-2958.2011.07758.x.

22. Rosch JW, Sublett J, Gao G, Wang YD, Tuomanen EI. 2008. Calcium effluxis essential for bacterial survival in the eukaryotic host. Mol Microbiol70:435– 444. https://doi.org/10.1111/j.1365-2958.2008.06425.x.

23. Rosch JW, Gao G, Ridout G, Wang YD, Tuomanen EI. 2009. Role of themanganese efflux system mntE for signalling and pathogenesis in Strep-tococcus pneumoniae. Mol Microbiol 72:12–25. https://doi.org/10.1111/j.1365-2958.2009.06638.x.

24. Shafeeq S, Kuipers OP, Kloosterman TG. 2013. The role of zinc in theinterplay between pathogenic streptococci and their hosts. Mol Micro-biol 88:1047–1057. https://doi.org/10.1111/mmi.12256.

25. Strausak D, Solioz M. 1997. CopY is a copper-inducible repressor of theEnterococcus hirae copper ATPases. J Biol Chem 272:8932– 8936. https://doi.org/10.1074/jbc.272.14.8932.

26. Smaldone GT, Helmann JD. 2007. CsoR regulates the copper effluxoperon copZA in Bacillus subtilis. Microbiology 153:4123– 4128. https://doi.org/10.1099/mic.0.2007/011742-0.

27. Corbett D, Schuler S, Glenn S, Andrew PW, Cavet JS, Roberts IS. 2011. Thecombined actions of the copper-responsive repressor CsoR and copper-metallochaperone CopZ modulate CopA-mediated copper efflux in theintracellular pathogen Listeria monocytogenes. Mol Microbiol 81:457– 472. https://doi.org/10.1111/j.1365-2958.2011.07705.x.

28. Vollmecke C, Drees SL, Reimann J, Albers SV, Lubben M. 2012. TheATPases CopA and CopB both contribute to copper resistance of thethermoacidophilic archaeon Sulfolobus solfataricus. Microbiology 158:1622–1633. https://doi.org/10.1099/mic.0.055905-0.

29. Jacobs AD, Chang FMJ, Morrison L, Dilger JM, Wysocki VH, Clemmer DE,Giedroc DP. 2015. Resolution of stepwise cooperativities of copperbinding by the homotetrameric copper-sensitive operon repressor(CsoR): impact on structure and stability. Angew Chem Int Ed Engl54:12795–12799. https://doi.org/10.1002/anie.201506349.

30. Johnson MD, Kehl-Fie TE, Klein R, Kelly J, Burnham C, Mann B, Rosch JW.2015. Role of copper efflux in pneumococcal pathogenesis and resis-tance to macrophage-mediated immune clearance. Infect Immun 83:1684 –1694. https://doi.org/10.1128/IAI.03015-14.

31. Singh K, Senadheera DB, Levesque CM, Cvitkovitch DG. 2015. ThecopYAZ operon functions in copper efflux, biofilm formation, genetictransformation, and stress tolerance in Streptococcus mutans. J Bacteriol197:2545–2557. https://doi.org/10.1128/JB.02433-14.

Updating an Operator Site for Copper Exporter Systems

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 13

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from

32. Fu Y, Tsui HC, Bruce KE, Sham LT, Higgins KA, Lisher JP, Kazmierczak KM,Maroney MJ, Dann CE, III, Winkler ME, Giedroc DP. 2013. A new structuralparadigm in copper resistance in Streptococcus pneumoniae. Nat ChemBiol 9:177–183. https://doi.org/10.1038/nchembio.1168.

33. Fu Y, Bruce KE, Wu HW, Giedroc DP. 2016. The S2 Cu(I) site in CupA fromStreptococcus pneumoniae is required for cellular copper resistance.Metallomics 8:61–70. https://doi.org/10.1039/c5mt00221d.

34. Neubert MJ, Dahlmann EA, Ambrose A, Johnson M. 2017. Copper chap-erone CupA and zinc control CopY Regulation of the pneumococcalcop operon. mSphere 2:e00372-17. https://doi.org/10.1128/mSphere.00372-17.

35. Glauninger H, Zhang Y, Higgins KA, Jacobs AD, Martin JE, Fu Y, Coyne RdHJ, Bruce KE, Maroney MJ, Clemmer DE, Capdevila DA, Giedroc DP. 2018.Metal-dependent allosteric activation and inhibition on the same mo-lecular scaffold: the copper sensor CopY from Streptococcus pneumoniae.Chem Sci 9:105–118. https://doi.org/10.1039/c7sc04396a.

36. Stoyanov JV, Brown NL. 2003. The Escherichia coli copper-responsivecopA promoter is activated by gold. J Biol Chem 278:1407–1410. https://doi.org/10.1074/jbc.C200580200.

37. Liu T, Ramesh A, Ma Z, Ward SK, Zhang LM, George GN, Talaat AM,Sacchettini JC, Giedroc DP. 2007. CsoR is a novel Mycobacterium tuber-culosis copper-sensing transcriptional regulator. Nat Chem Biol 3:60 – 68.https://doi.org/10.1038/nchembio844.

38. Changela A, Chen K, Xue Y, Holschen J, Outten CE, O’Halloran TV,Mondragón A. 2003. Molecular basis of metal-ion selectivity and zepto-molar sensitivity by CueR. Science 301:1383–1387. https://doi.org/10.1126/science.1085950.

39. Arguello JM, Gonzalez-Guerrero M, Raimunda D. 2011. Bacterial transi-tion metal P(1B) ATPases: transport mechanism and roles in virulence.Biochemistry 50:9940 –9949. https://doi.org/10.1021/bi201418k.

40. Hava DL, Camilli A. 2002. Large-scale identification of serotype 4 Strep-tococcus pneumoniae virulence factors. Mol Microbiol 45:1389 –1406.

41. Safo MK, Zhao QX, Ko TP, Musayev FN, Robinson H, Scarsdale N, WangAHJ, Archer GL. 2005. Crystal structures of the BlaI repressor fromStaphylococcus aureus and its complex with DNA: insights into transcrip-tional regulation of the bla and mec operons. J Bacteriol 187:1833–1844.https://doi.org/10.1128/JB.187.5.1833-1844.2005.

42. Arede P, Oliveira DC. 2013. Proteolysis of mecA repressor is essential forexpression of methicillin resistance by Staphylococcus aureus. Antimi-crob Agents Chemother 57:2001–2002. https://doi.org/10.1128/AAC.02510-12.

43. Portmann R, Magnani D, Stoyanov JV, Schmechel A, Multhaup G, SoliozM. 2004. Interaction kinetics of the copper-responsive CopY repressorwith the cop promoter of Enterococcus hirae. J Biol Inorg Chem9:396 – 402. https://doi.org/10.1007/s00775-004-0536-1.

44. Magnani D, Barre O, Gerber SD, Solioz M. 2008. Characterization of the

CopR regulon of Lactococcus lactis IL1403. J Bacteriol 190:536 –545.https://doi.org/10.1128/JB.01481-07.

45. Collins TC, Dameron CT. 2012. Dissecting the dimerization motif ofEnterococcus hirae’s Zn(II)CopY. J Biol Inorg Chem 17:1063–1070. https://doi.org/10.1007/s00775-012-0919-7.

46. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic localalignment search tool. J Mol Biol 215:403– 410. https://doi.org/10.1016/S0022-2836(05)80360-2.

47. Ciesielski GL, Hytonen VP, Kaguni LS. 2016. Biolayer interferometry: anovel method to elucidate protein-protein and protein-DNA interactionsin the mitochondrial DNA replisome. Methods Mol Biol 1351:223–231.https://doi.org/10.1007/978-1-4939-3040-1_17.

48. Mrazek J, Xie S. 2006. Pattern locator: a new tool for finding localsequence patterns in genomic DNA sequences. Bioinformatics 22:3099 –3100. https://doi.org/10.1093/bioinformatics/btl551.

49. Gupta RS, Lo B, Son J. 2018. Phylogenomics and comparative genomicstudies robustly support division of the genus Mycobacterium into anemended genus Mycobacterium and four novel genera. Front Microbiol9:67. https://doi.org/10.3389/fmicb.2018.00067.

50. Bailey TL, Johnson J, Grant CE, Noble WS. 2015. The MEME Suite. NucleicAcids Res 43:W39 –W49. https://doi.org/10.1093/nar/gkv416.

51. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, Ren J, Li WW,Noble WS. 2009. MEME Suite: tools for motif discovery and searching.Nucleic Acids Res 37:W202–W208. https://doi.org/10.1093/nar/gkp335.

52. Bailey TL, Elkan C. 1994. Fitting a mixture model by expectation maxi-mization to discover motifs in biopolymers. Proc Int Conf Intell Syst MolBiol 2:28 –36.

53. Odermatt A, Solioz M. 1995. Two trans-acting metalloregulatory proteinscontrolling expression of the copper-ATPases of Enterococcus hirae. JBiol Chem 270:4349 – 4354. https://doi.org/10.1074/jbc.270.9.4349.

54. Dougherty WG, Carrington JC, Cary SM, Parks TD. 1988. Biochemical andmutational analysis of a plant virus polyprotein cleavage site. EMBO J7:1281–1287.

55. Slager J, Aprianto R, Veening JW. 2018. Deep genome annotation of theopportunistic human pathogen Streptococcus pneumoniae D39. NucleicAcids Res 46:9971–9989. https://doi.org/10.1093/nar/gky725.

56. Gregory PD, Lewis RA, Curnock SP, Dyke KG. 1997. Studies of therepressor (BlaI) of beta-lactamase synthesis in Staphylococcus aureus.Mol Microbiol 24:1025–1037. https://doi.org/10.1046/j.1365-2958.1997.4051770.x.

57. van Opijnen T, Camilli A. 2012. A fine scale phenotype-genotype viru-lence map of a bacterial pathogen. Genome Res 22:2541–2551. https://doi.org/10.1101/gr.137430.112.

58. Stols L, Gu MY, Dieckman L, Raffen R, Collart FR, Donnelly MI. 2002. Anew vector for high-throughput, ligation-independent cloning encodinga tobacco etch virus protease cleavage site. Protein Expr Purif 25:8 –15.https://doi.org/10.1006/prep.2001.1603.

O’Brien et al.

May/June 2020 Volume 5 Issue 3 e00411-20 msphere.asm.org 14

on October 14, 2020 by guest

http://msphere.asm

.org/D

ownloaded from